Convert observer data models to Pydantic BaseModel with timestamps

Switch ProcessorStartupTiming, StartupTimingReport, and
TransportTimingReport from dataclasses to Pydantic BaseModel. Add
start_time (Unix timestamp) fields and wall clock conversion for
monotonic observer timestamps.
This commit is contained in:
Mark Backman
2026-03-02 11:01:26 -05:00
parent 68e8732e72
commit 75669b12a2
2 changed files with 32 additions and 8 deletions

View File

@@ -34,9 +34,11 @@ Example::
task = PipelineTask(pipeline, observers=[observer])
"""
from dataclasses import dataclass, field
import time
from typing import Dict, List, Optional, Tuple, Type
from pydantic import BaseModel, Field
from pipecat.frames.frames import BotConnectedFrame, ClientConnectedFrame, StartFrame
from pipecat.observers.base_observer import BaseObserver, FrameProcessed, FramePushed
from pipecat.pipeline.base_pipeline import BasePipeline
@@ -47,42 +49,45 @@ from pipecat.processors.frame_processor import FrameProcessor
_INTERNAL_TYPES = (PipelineSink, PipelineSource, BasePipeline)
@dataclass
class ProcessorStartupTiming:
class ProcessorStartupTiming(BaseModel):
"""Startup timing for a single processor.
Parameters:
processor_name: The name of the processor.
start_time: Unix timestamp when the processor's start() began.
duration_secs: How long the processor's start() took, in seconds.
"""
processor_name: str
start_time: float
duration_secs: float
@dataclass
class StartupTimingReport:
class StartupTimingReport(BaseModel):
"""Report of startup timings for all measured processors.
Parameters:
start_time: Unix timestamp when the first processor began starting.
total_duration_secs: Total wall-clock time from first to last processor start.
processor_timings: Per-processor timing data, in pipeline order.
"""
start_time: float
total_duration_secs: float
processor_timings: List[ProcessorStartupTiming] = field(default_factory=list)
processor_timings: List[ProcessorStartupTiming] = Field(default_factory=list)
@dataclass
class TransportTimingReport:
class TransportTimingReport(BaseModel):
"""Time from pipeline start to transport connection milestones.
Parameters:
start_time: Unix timestamp of the StartFrame (pipeline start).
bot_connected_secs: Seconds from StartFrame to first BotConnectedFrame
(only set for SFU transports).
client_connected_secs: Seconds from StartFrame to first ClientConnectedFrame.
"""
start_time: float
bot_connected_secs: Optional[float] = None
client_connected_secs: Optional[float] = None
@@ -176,9 +181,19 @@ class StartupTimingObserver(BaseObserver):
# Bot connected timing (stored for inclusion in the transport report).
self._bot_connected_secs: Optional[float] = None
# Wall clock reference for converting monotonic ns to Unix timestamps.
self._wall_clock_ref: Optional[float] = None
self._mono_clock_ref_ns: Optional[int] = None
self._register_event_handler("on_startup_timing_report")
self._register_event_handler("on_transport_timing_report")
def _mono_to_wall(self, mono_ns: int) -> float:
"""Convert a monotonic nanosecond timestamp to a Unix wall clock time."""
if self._wall_clock_ref is None or self._mono_clock_ref_ns is None:
return 0.0
return self._wall_clock_ref + (mono_ns - self._mono_clock_ref_ns) / 1e9
def _should_track(self, processor: FrameProcessor) -> bool:
"""Check if a processor should be tracked for timing.
@@ -212,6 +227,8 @@ class StartupTimingObserver(BaseObserver):
if self._start_frame_id is None:
self._start_frame_id = data.frame.id
self._start_frame_arrival_ns = data.timestamp
self._wall_clock_ref = time.time()
self._mono_clock_ref_ns = data.timestamp
elif data.frame.id != self._start_frame_id:
return
@@ -263,6 +280,7 @@ class StartupTimingObserver(BaseObserver):
self._timings.append(
ProcessorStartupTiming(
processor_name=processor.name,
start_time=self._mono_to_wall(arrival_ts),
duration_secs=duration_secs,
)
)
@@ -284,6 +302,7 @@ class StartupTimingObserver(BaseObserver):
delta_ns = data.timestamp - self._start_frame_arrival_ns
client_connected_secs = delta_ns / 1e9
report = TransportTimingReport(
start_time=self._mono_to_wall(self._start_frame_arrival_ns),
bot_connected_secs=self._bot_connected_secs,
client_connected_secs=client_connected_secs,
)
@@ -296,8 +315,10 @@ class StartupTimingObserver(BaseObserver):
self._startup_timing_reported = True
total = sum(t.duration_secs for t in self._timings)
start_time = self._timings[0].start_time if self._timings else 0.0
report = StartupTimingReport(
start_time=start_time,
total_duration_secs=total,
processor_timings=self._timings,
)

View File

@@ -151,9 +151,11 @@ class TestStartupTimingObserver(unittest.IsolatedAsyncioTestCase):
report = reports[0]
self.assertIsInstance(report, StartupTimingReport)
self.assertIsInstance(report.total_duration_secs, float)
self.assertGreater(report.start_time, 0)
for timing in report.processor_timings:
self.assertIsInstance(timing.processor_name, str)
self.assertIsInstance(timing.duration_secs, float)
self.assertGreater(timing.start_time, 0)
async def test_excludes_internal_processors(self):
"""Test that internal pipeline processors are excluded by default."""
@@ -211,6 +213,7 @@ class TestStartupTimingObserver(unittest.IsolatedAsyncioTestCase):
self.assertEqual(len(transport_reports), 1)
report = transport_reports[0]
self.assertIsInstance(report, TransportTimingReport)
self.assertGreater(report.start_time, 0)
self.assertGreater(report.client_connected_secs, 0)
self.assertIsNone(report.bot_connected_secs)